Hi all,
Sorry for asking this question again, but I haven’t
found any answers even in the maths books…
I am doing a VBM analysis comparing 2 groups.
Unfortunately, group 1 is significantly older than
group 2.
When NOT using age as covariate there are
significantly gray differences between these two groups (which also make
sense).
But the factor group also correlates significantly
with age (as expected)
Using age as a single covariate the significant
differences in the gray matter disappear.
When using the covariate option interaction and
centering around factor1 , the significant differences are back again.
Is it allowed to use this option even if I am only
looking at the group differences (there are now 2 covariate regressors, one for
each group)?
Many thanks
boris
Priv.-Doz. Dr. Boris
Suchan
Institute of Cognitive
Neuroscience
Ruhr University Bochum
Universitätsstr. 150
44780 Bochum
GAFO 05/613
Tel.: + 49 234 3227575
Fax: + 49 234 3214622
http://www.ruhr-uni-bochum.de/neuropsy/mitarbeiter/boris_suchan.html